HIV-accelerated Liver Disease in persons of African

Report
High Risk of Liver Fibrosis and Cirrhosis
Among HIV/HBV Co-Infected Persons in
Rakai, Uganda
Lara Stabinski1, Gregory D. Kirk2, Steve J Reynolds1, Ponsiano
Ocama3, Francis Bbosa4, Melissa Saulynas2, V. Kiggundu4, Dave
Thomas2, Ron Gray2, Tom Quinn1 & Chloe Thio2
(1) NIH/LIR Bethesda, Maryland, and the NIH ICER Rakai, Uganda (2) Johns
Hopkins University, Baltimore, Maryland (3) Makerere University, Kampala,
Uganda, (4) Rakai Health Sciences Program, Rakai, Uganda
Background
• Liver disease is a leading cause of death among HIV-infected
persons in western cohorts, especially among those coinfected with hepatitis B or C viruses (HBV, HCV)
• Data regarding the effects of HIV, hepatitis B and HAART on
liver disease in Africa remain extremely sparse
• Estimates of hepatotoxicity among HIV-infected persons
based on liver enzyme elevation are low in Africa
• Evidence that liver enzyme elevation may have substantial
limitations as a marker of liver disease in HIV + persons
•
Biopsy studies to ascertain liver disease are invasive and
often difficult to conduct in resource-limited settings
Methods: A non-invasive, cross
sectional study in Uganda
• Transient elastography (TE)
(FibroScan®) used to estimate
fibrosis
• Population prevalence of
HIV/HBV co-infection 5%
• Participants
– 61 HBV/HIV co-infected
Rakai Health Sciences
Program (RHSP) ART
program
– 51HBV mono-infected RHSP
population cohort
– All included participants had
valid TE scans, available HBV
DNA
Methods
• After consent, participants underwent a
structured interview, collection of biological samples,
and transient elastography to obtain liver stiffness
measurements (LSM) for quantitation of liver fibrosis.
• LSM cutoffs (in kilopascals, kPa)
– Significant fibrosis (equivalent to Metavir F ≥2, LSM ≥9.3)
– Cirrhosis (Metavir F>4, LSM ≥12.3).
• Correlates of liver fibrosis were identified using modified
Poisson regression to estimate prevalence rate ratios
(PRR) with 95% confidence intervals (CI).
Results: Demographics
HIV/HBV
Co-infected
HBV
Mono-infected
27 (44)
24 (47)
37 (32-57)
30 (25-46)
Use Liquor
12 (20)
5 (10)
On HAART
33 (54)
-
21 (11-48)
-
406 (179-1152)
-
22(36)
-
Male Gender
†Median Age (IQR)
Median HAART Duration,
months (IQR)
Median CD4 count (cells/mm3)
Nadir CD4 <200
Data represent N (%) or median (interquartile range [IQR])
† p<0.05 HIV/HBV co-infected v. HBV mono-infected
Results: ALT & HBV DNA Levels
HIV/HBV
Co-infected
HBV
Mono-infected
25 (12-71)
24(13-37)
54 (89)
50 (98)
Grade1 (1.25-2.5 X ULN)
5 (8)
1 (2)
Grade 2 (2.5 -5X ULN)
1 (2)
-
Grade 3 (5-10X ULN)
1 (2)
-
Grade 4 (>=10X ULN)
-
-
35 (57)‡
26 (51)
Median ALT (IU/L)
ACTG ALT Hepatotoxicity Criteria (%)
Grade 0 (0-1.25X ULN)
HBV DNA<100 IU/ml (below detection)
‡ HBV DNA <100 in 60% of co-infected on HAART vs. 20% in those not yet
on HAART (p=0.002)
Prevalence of Liver Disease
HIV/HBV Co-infected
Significant
Fibrosis
28%
HBV Mono-infected
Significant
Fibrosis
14%
Cirrhosis
14%
Fibrosis
41%
Cirrhosis
59%
Fibrosis
86%
Predictors of Significant Fibrosis in
Overall Study Population
Gender, age, liquor use, and HIV/HBV co-infection
Univariate PRR
P
Multivariate
PRR
p
Male Gender
1.2 (0.6-2.4)
0.622
1.0 (0.4-2.3)
0.995
Age (per year)
1.0 (1.0-1.0)
0.448
1.0 (0.9-1.0)
0.095
Use liquor
1.9 (0.9-4.0)
0.113
1.9 (0.7-1.9)
0.202
HIV /HBV vs. HBV only
2.0 (0.9-4.5)
0.083
2.3 (1.0-5.0)
0.041
Predictors of Significant Fibrosis in
Overall Study Population
HAART and nadir CD4 in HIV/HBV co-infected compared to HBV mono-infected
Multivariate ‡
PRR
p
HBV mono-infected
1 (ref)
-
HIV/HBV, No HAART
nadir CD4>=200
2.7 (1.2-6.5)
0.022
HIV/HBV, No HAART
nadir CD4< 200
5.0 (1.8-13.9)
0.002
HIV/HBV, On HAART
nadir CD4>=200
1.3 (0.3-5.2)
0.746
HIV/HBV, On HAART
nadir CD< 200
1.8 (0.6-5.6)
0.299
‡ adjusted for age, gender and liquor use
Predictors of Significant Fibrosis in
Overall Study Population
HBV DNA accounting for HIV/HAART status
Multivariate
PRR‡
p
†HBV DNA >100 IU/ml
2.6 (1.0-6.5)
0.040
HIV positive, No HAART
2.4 (1.0-5.3)
0.039
HIV positive on HAART
1.8 (0.7-4.7)
0.257
† Numbers of participants were insufficient to further characterize risks
with other categorizations of HBV DNA above 100 IU/ml
‡ adjusted for age, gender and liquor use
Notable Predictors of Liver Disease
in HIV/HBV co-infected Population
•HAART associated with a 60% reduction in fibrosis; PRR
0.4 (95% CI 0.1-1.0), adjusted for age, gender and nadir
CD4 <200 cells/mm3
• Nadir CD4 <200 cells/mm3 not significantly associated
with fibrosis; PR 1.7 (95% CI 0.7-4.2)
• HBV DNA >100 IU/ml associated with an increased risk
of fibrosis; PRR 3.0 (95% CI 1.0-9.3), controlled for age,
gender, HAART and nadir CD4 <200
Conclusion
• In HIV/HBV co-infected persons, the prevalence
of significant fibrosis is double that of HBV mono-infected
persons
•
HAART appears to provide protection against liver fibrosis
among HIV-infected persons; early initiation of HAART
may be useful in co-infection in resource limited settings
• These data underscore the need for effective treatment for
HBV in resource-limited settings as HAART is scaled up
Acknowledgements
Rakai Health Sciences Program
– Maria Wawer
– Ron Gray
– Anthony Ndyanabo
– Francis Bbosa
– Victor Ssempijja
– Denis Ssenyondwa
– Gladys Namuyaba
– Violet Nkalubo
– Grace Kigozi
– Valerian Kiggundu
– Gertrude Nakigozi
– Iga Boaz & the RHSP Lab team
– Fred Nalugoda
– Tom Lutalo
– Godfrey Kigozi
– Joseph Kagaayi
– David Serwadda
•
NIH/LIR & the NIH Uganda ICER
–
–
–
–
–
–
Tom Quinn
Steven J. Reynolds
Oliver Laeyendecker
Andrew Redd
Aaron Tobian
Kevin Newell (SAIC, INC)
Makerere University, Uganda
– Ponsiano Ocama
– Kenneth Opio
– Emmanuel Seremba
– Godfrey Gemagaine
Johns Hopkins University
– Gregory D. Kirk
– Dave Thomas
– Chloe Thio

similar documents